RFID Anti-collision Algorithm

RFID (Radio Frequency Identification) is an important technology of the Internet of Thingstechnology. With the rapid development of the Internet of Things, RFID technology has got greatdevelopment. RFID transmits information through radio frequency signal and recognizes theobjects without direct contact. It is a kind of automatic identification technology. Compared withother automatic identification technology that has been widely used at present, the advantages ofRFID technology is significant. So the technology has a great space for development in thefuture and has a great application potential.In this thesis several kinds of anti-collision algorithms in common is deeply studied, such aspure ALOHA, slotted ALOHA algorithm, frame slotted ALOHA algorithm, dynamic frameslotted ALOHA algorithm that base on ALOHA mechanism and binary search algorithm,backward binary search algorithm, dynamic binary search algorithm, jump dynamic binarysearch algorithm that based on binary tree. On the analysis of the core idea of the algorithm andtheir respective advantages and disadvantages, aiming to improve the system throughput andreduce the search times and communication, two improved algorithms are proposed based onALOHAalgorithm and binary search algorithm accordingly.The main work is as follows:(1) Based on the study of RFID system composition and working principle, the strategy tosolve the collision problem is deep studied. Because of the ALOHA algorithm is not instable, therecognition efficiency is not high. When there are a large number of electronic tags, some ofthem can’t be recognized in a fairly long period of time and will be "starve to death". Animproved algorithm is proposed. When there are a large number of electronic tags, the proposedalgorithm divides the electronic tags into two groups by setting a threshold. Then only one groupthat meet the rule response the reader. The proposed algorithm restricts the response electronictag quantity by grouping to achieve high recognition efficiency. The algorithm can effectivelysolve the sharp decline in the efficiency and response time problem when there are a largenumber of electronic tags.(2) In order to solve the problem that the binary tree algorithm communication is big andtakes long time to recognize the tags, an improved algorithm is proposed in this thesis. Based onthe binary tree algorithm, the algorithm introduces dynamic adjustment algorithm. The electronictags are divided into two groups by embedding a counter in the electronic tag chip. Aftergrouping, only one group can be recognized each time. Because the each group tag contains1oreven or odd number, the tags can be recognized directly when there are only two collision bit. Sothe search time can be reduced. We let the stack store the data, and let the reader send commandparameters associated with the highest collision only, which can significantly reduce the amountof communication between the reader and the tag.